Awesome

docker-jupyter-dl-gpu

GPU-enabled docker image for deep learning.

The image provides a complete environment for deep learning in pytorch, including ipython, conda, cuda, pytorch, torchvision, tensorboard, numpy, pandas, scipy, matplotlib, seaborn, and many more (see below for a full package list).

This image is fully compatible as the user image for a JupyterHub deployment.

Prerequisites

• docker
• (optional) nvidia-docker - required to run on GPU

Usage

Jupyter:

docker run --rm -p 8888:8888 -v <local path>:/home/jovyan/ timesler/jupyter-dl-gpu:latest

Jupyter lab:

docker run --rm -p 8888:8888 -v <local path>:/home/jovyan/ timesler/jupyter-dl-gpu:latest jupyter lab

IPython:

docker run --rm -p 8888:8888 -v <local path>:/home/jovyan/ timesler/jupyter-dl-gpu:latest ipython

Bash:

docker run --rm -p 8888:8888 -v <local path>:/home/jovyan/ timesler/jupyter-dl-gpu:latest bash

Depending on your system's configuration, you may need to call nvidia-docker instead of docker or specify nvidia as the docker runtime.

Complete package list

Below is a complete list of conda packages installed in the image: